HAART has dramatically changed the HIV epidemic, delaying disease, prolonging life and altering the nature of HIV-associated neurocognitive disorders (HAND) from overt dementia to cognitive/motor disorders. Antiretroviral drugs have differing capacities to penetrate into the CNS. It is not clear from HAART therapy in humans whether CNS-penetrant HAART improves CNS outcomes by reducing viral replication and inflammation in the CNS or paradoxically contributes to long term CNS damage due to higher CNS levels of potentially neurotoxic drugs. We developed and characterized a rigorous SIV macaque model of combined antiretroviral therapy (cART) that is very similar to HAART in HIV-infected individuals, combining classes of antiretroviral drugs that have low CNS penetrance. The value of this model is: 1) Both CD4+ T cells and monocyte/macrophages are infected, 2) tissues including brain and spleen harbor virus in HAART-treated macaques, accurately modeling HAART in HIV-infected humans, 3) the number of latently infected resting CD4+ cells in blood and lymphoid tissues is comparable to that in HIV-infected patients on HAART. These advantages demonstrate the model's ability to study latent viral reservoirs in brain and the peripheral nervous system. Virus replication in brain is detectable at low levels using a sensitive single copy assay; SIV DNA levels were comparable to untreated SIV-infected macaques, and there were ongoing inflammatory changes in brain. We now propose to study a SIV model using CNS-penetrant cART (pcART) to determine its impact and compare it directly with non-CNS-penetrant ART (ncART) for ability to 1) control viral replication& inflammation 2) reduce viral DNA reservoirs in brain 3) preserve markers of neuronal function. We will also examine the relative ability of pcART versus ncART to prolong time to virus reactivation after stopping cART. Our hypothesis is that CNS-penetrant cART will be more effective than non-penetrant cART in reducing residual viral replication in CSF and CNS, in reducing viral reservoirs in brain, and potentially other tissues due to increased penetration, as measured by viral DNA and residual virus replication, in reducing CNS inflammation, and in delaying virus reactivation upon withdrawal of cART. However, CNS-penetrant cART might also have neurotoxic effects with increased neuronal damage in both CNS and PNS. We will: 1) Characterize and compare decay kinetics of SIV in plasma and CSF using an optimized CNS-penetrant cART (pcART) to non- penetrant cART (ncART) in our SIV macaque model; quantitate and compare residual virus replication in plasma, PBMCs, CSF and tissues in pcART versus ncART-treated animals; 2) Examine viral latency in CD4+T cells and monocytes in blood and CD4+T cells and macrophages in tissues of SIV-infected macaques treated with pcART & ncART; 3) Measure CNS and PNS markers of neuronal function in SIV-infected macaques with pcART & ncART regimens to evaluate neuroprotection and neurotoxicity; 4) Compare ability of pcART versus ncART to delay or prevent reactivation of virus from reservoirs including the CNS in SIV-infected macaques.